Statistical hypothesis testing in case-control studies on <i>HLA</i> and disease associations

DOI Web Site
  • Ohashi Jun
    Laboratory of Human Genome Diversity, Department of Biological Sciences, Graduate School of Science, The University of Tokyo

Bibliographic Information

Other Title
  • <i>HLA</i>と疾患の関連研究で用いる統計学的仮説検定

Search this article

Abstract

<p>A number of HLA alleles have been reported to be associated with various traits such as autoimmune diseases, infectious diseases, and adverse drug reaction in case-control association studies. Since there are many alleles at classical class I and class II HLA loci, a stringent significance level as corrected by the Bonferron method has been used in the statistical hypothesis testing to reduce false positives. However, the Bonferroni correction is too conservative to detect HLA alleles showing weak association, and thus many false negatives seem to have occurred in HLA-disease association studies. Here, I describe a chi-square test based on 2 by 2 contingency table for evaluating the association of each HLA allele at the locus with a dichotomous trait such as disease. In addition, I review the adjustment of significance level in association tests for loci with many alleles.</p>

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top